On average, the insurance industry delivers only mediocre returns. At the same time, it is becoming increasingly polarized: at one extreme are a few steady winners that achieve consistently high returns; at the other, a majority that deliver only average or worse results. Industry pundits point to the winners and recite a now-familiar litany: superior marketing, better risk selection and underwriting, and top-quality claim performance. To be sure, these factors do play an important role, and they certainly separate winners such as Progressive and AIG from the rest.

Yet few observers even mention investment management as a skill that is critical to sustained success. Indeed, when senior insurance executives talk about their top priorities, improving investment performance seldom makes the list. But have they got it right? Is investment management less important than other functions?

Consider the following:

• From 1990 to 1994, US property and casualty investment income as a percentage of written net premiums was 19.1 percent, compared with an insurance underwriting result of -9.9 percent.
• Leading performers in insurance investment management achieved considerably higher investment results than average players, with, for example, 1.3 percentage point higher returns in the United States and 1.4 in Germany. In the United States, this translated into a 4 percentage point higher return on equity than the industry average.
• Some of the most notable losses in the industry have been related to investment: for instance, Equitable’s real estate and GIC losses, Colonia’s derivative losses, and First Executive’s junk-bond losses.

Given these facts, why isn’t more attention being paid to the investment function? Part of the blame must go to several persistent myths:

• "Nobody has sustained better-than-average performance." While this may be true in the mutual fund and pension industry, it is decidedly not true in insurance. As publicly available data show, some insurers—SAFECO in the United States, Zurich in Switzerland, and Victoria in Germany, for example—outperform their peers year after year.
• "It’s too hard to make a difference in investment management performance." False. While improving performance isn’t easy, the leverage in investment management is at least as great as in other insurance functions. To create $1 billion in shareholder value, for example, a $10 to $15 billion premium insurer would need either to improve investment returns by 30 basis points (bp) or to expand volume by approximately 6 percent (Exhibit 1). Which is more difficult?
  
• "We’re in the insurance business, not the investment management business." Wrong. The $2.5 trillion in insurance investment assets in the United States constitutes one-third of the country’s entire institutional assets. Moreover, assets managed by insurance companies account for 40 percent of the assets of the world’s top 100 asset managers. Insurers are in the investment management business.
• "Whatever we do, competitors will follow and copy us." Maybe, but so what? Unlike many other actions, improving investment performance is not a zero-sum game. In addition, competitors will not easily discern what actions an insurer is taking; only the results will be visible.

We believe that the economic opportunity for insurers to improve their investment performance is considerable: 100 bp or even more, depending on the situation. Here, we examine the obstacles that stand in the way of superior investment performance, and outline five levers for improving performance. The importance of each lever to an insurer will depend on its starting point and the nature of the underlying insurance business.

Obstacles to improved investment performance

Even when insurance companies are aware of the opportunity presented by superior investment management, many obstacles can prevent them becoming first-rate global investors. These obstacles fall into three categories:

• The need for investment profiles to reflect the pattern of cash outflows so as to minimize exposure
• The organizational structure, legal entity structure, and need for local focus that prevent insurers developing a macro-view of all their investment activities
• Typical insurance performance measurement systems, which often obscure real investment performance.

Investment profiles

Insurance investment activities have traditionally been viewed as ancillary to underwriting. Put another way, insurance is the business of generating liabilities that must be matched by investment assets. Consequently, much actuarial time and attention is spent forecasting liabilities and payout patterns. Then a "fudge" factor may be added for unexpected catastrophic losses. This translates into an asset mix that is highly, if not excessively, liquid and fixed income in nature, rather than a portfolio mix that is optimal from an investor’s perspective.

For global players, the country asset mix usually mirrors the underlying insurer liability mix, often regardless of whether the individual asset classes concerned represent attractive investment opportunities from a timing, diversification, or fundamental perspective. The asset portfolio is driven almost exclusively by considerations of balance-sheet management rather than by an investment viewpoint. If individual investors behaved in this way, they would be castigated as recklessly conservative.

Organization structure and local focus

Most insurers, even purely domestic ones, consist of numerous local companies, legal entities, or business units. Such structures usually come about through acquisition, a desire to achieve focus, a need to have several companies for filing purposes, or simply a historic pattern of growth.

In turn, investment activities, and particularly asset allocation, are often driven by an underwriting-related organizational model. A typical multinational insurer will have investment activities in each of the countries in which it operates. Asset allocation, diversification, and even hedging decisions are made at local level, often with scant concern for the way in which the overall entity is allocated or invested. The result is not only reckless conservatism but:

• Systematic underinvestment. The sum total of all the individual asset allocation decisions at local level is usually far more conservative and less diversified than an optimal asset allocation for the combined entity would be. When each unit makes a set of conservative assumptions and builds in extra safety margins that are aggregated at the enterprise level, a highly inefficient asset mix ensues.
• Excessive costs. Having multiple managers manage multiple portfolios of similar assets and asset classes is costly, given the high fixed costs of investment management. Similarly, managing multiple business unit portfolios individually incurs higher hedging costs because it is difficult to take advantage of any natural hedges that exist within the combined entity and because purchasing hedges for small portfolios is relatively expensive.
• Inability to achieve economies of skill. Investment talent is scarce and costly. When a number of investment managers in separate business units are doing essentially the same job, the outcome may be high costs, mediocre skills, or both.

Performance measurement

Most insurance companies’ performance measurement systems reward business unit managers for total results—for the sum of the combined ratio and investment ratio. As a result, investment decisions are often taken by business unit managers, even if these managers lack skill in investment management, and even if the decisions they take are suboptimal from the perspective of the company as a whole.

At one carrier, a business unit head overruled asset allocation decisions in favor of a fixed-income portfolio. The reason? A desire for more current yield—albeit at the cost of lower total returns. Situations like this are magnified in global companies, where currency translations can distort a business unit’s economics and lead to gamesmanship with the investment portfolio.

A more serious consequence of this approach is that fundamental risk management decisions are made by default. Insurers take three kinds of risk: underwriting risk (for example, choice of business, pricing); leverage risk (that is, premiums to surplus); and investment risk (for example, choice of assets). Where and how to deploy risk capital needs to be an explicit, measured decision driven by market conditions and an understanding of the options available. Yet such decisions are made implicitly or not at all under the standard insurance performance measurement approach.

Why some insurance companies continue to hold what are essentially underwriting unit managers accountable for investment results is unclear. Most banks have long since abandoned this practice. Instead, they measure the performance of a lending unit by gross interest income less a matched opportunity rate for the cost of funds. Left to their own devices, savvy lending unit heads would slash their cost of funds (probably by taking short-term deposits), even if they are lending "long." Banks prevent this by uncoupling the asset-generating and deposit-gathering functions and measuring them separately, rather than on a net basis. Transfer pricing is used to measure overall unit profitability.

Can insurers act more like banks? Yes, if they are willing to redesign the investment process and restructure the investment function. While such measures are not radical, they are substantial.

Five levers for improving performance

Five levers are available to senior management as it considers how to improve investment performance. These levers include changes both to what gets done and to how it gets done. They are:

• Strategic asset allocation
• Tactical asset allocation
• Security selection
• Risk management
• Efficiency and effectiveness of the investment function.

The first three levers entail redesigning the investment process; the fourth and fifth involve restructuring the investment function.

Strategic asset allocation

Strategic asset allocation is about determining the optimal long-term risk/return profile of an investment portfolio. A well-conceived strategic asset allocation process can improve risk-adjusted performance in a couple of critical ways:

• Optimizing the risk/return profile with respect to the company’s appetite for risk and capital structure. This profile will change over time as the insurance company trades off risk in the investment portfolio against underwriting risk (for instance, pricing or classes of business) and operating leverage in the business (premium to surplus ratio).
• Capturing enterprise-wide opportunities, particularly international opportunities, through a globally optimal asset allocation.

By comprehensively evaluating risk appetite, liability structure, and internal and external constraints, an insurer can achieve an optimal long-term strategic asset allocation among various classes of assets.

Improving strategic asset allocation can yield impressive results. We have seen potential gains ranging from 30 to 100 bp, depending on the current level of asset allocation optimization within a company and the extent to which international investment opportunities can be exploited. The scale of the potential is hardly surprising, given that the largest share of investment return in an average portfolio typically derives from asset allocation.

How does a company actually set about achieving optimal strategic asset allocation? There are three steps:

1. Identify and evaluate all current internal and external constraints on investment portfolios. These include risk-bearing capital, local regulations, liability structure, and any constraints arising from business unit autonomy.

2. Determine the theoretically optimal asset allocation for the entire company. This involves modeling an efficient frontier for the company and calculating the optimal asset mix on the basis of return and risk (Exhibit 2). Efficient frontier methodology consists of maximizing a portfolio’s return for a given risk by reallocating the weighting of each asset class on the basis of its return expectation and its covariance with all other asset classes. This frontier will differ from company to company, and even from business unit to business unit, according to each unit’s ability to invest in certain asset classes. Other analytical approaches can also be used to arrive at optimal strategic asset allocation. Asset/liability modeling, for example, may be appropriate for institutions that can predict liability streams accurately.

3. Overlay the current constraints identified in step 1 on the optimal strategic asset allocation from step 2 in order to identify gaps. Next, the constraints identified in step 1 must be evaluated in terms of the shortfall in investment performance relative to what would be achieved under optimal asset allocation. For the largest opportunities, a company should then ask: how can we relax the constraint? Typical constraints include:

• Underinvestment at an international level, often as a result of local regulations or local business unit decisions
• Investment in covariant asset classes, which impairs the desired diversification
• Underuse of tax minimization tools, such as Bermuda facilities.

Most constraints can be overcome, but usually only by taking an enterprise-wide investment perspective. This may come at the expense of local unit autonomy or through a reduction in the ability to measure business units on both insurance and investment results. For example, although a fixed-income yen investment would show better current income for a Japanese business unit, it might make more sense in terms of strategic asset allocation to invest a portion of the portfolio in Thai equities. For every situation of this kind, the company must determine whether to pursue the opportunity that arises.

Tactical asset allocation

Given an optimal strategic asset allocation, the second lever is tactical asset allocation: the deviation from long-term asset allocation to capture medium-term market opportunities.

Essentially, tactical asset allocation is a matter of market timing and adding value by adjusting the strategic asset allocation in accordance with superior insight. It is about active rather than passive investment management. Passive management means adhering exclusively to the strategic asset allocation; active management means departing from it. Effective tactical asset allocation can offer an incremental economic opportunity of approximately 30 bp.

How far tactical asset allocation should be pursued depends on a company’s skills and ability to measure the effects of tactical asset allocation decisions. The skill point is straightforward. Questions to ask include:

• How good are our overall market timing skills?
• Do we have strengths in certain asset classes or categories?
• Are these skills spread across the organization or concentrated in units?

Measurement is a more subtle issue. The overall performance of an asset class does not necessarily indicate whether a tactical decision was sound. In fact, poor tactical decisions can be associated with high-performing asset classes when an overall strategic allocation was sound and the asset class performed well. The tactical asset allocation must be tracked against benchmarks tailored to the strategic asset allocation to determine what value has been added or subtracted by tactical asset decisions.

Security selection

After tactical asset allocation decisions about which markets and asset classes to invest in have been made, security selection presents an opportunity to generate additional returns by identifying short- and medium-term opportunities at the security level. Improvements here may yield up to an extra 20 bp.

Security selection is the ultimate active/passive tradeoff. Key management questions include:

• Can we capture short- and medium-term opportunities at the security level to maximize our risk-adjusted investment performance ...

  ... by leveraging our institutional security selection skills?

  ... by leveraging our access to information?

  ... by outsourcing certain investments (such as passive management or exotic asset classes for which in-house skills are not cost-effective)?

• Do we understand the drivers of our current over- or underperformance at security selection level?
• Do we have the skills and systems to measure and monitor performance?

The answers to these questions will determine the extent to which an insurer can profitably pursue security selection.

Risk management

Risk management is the ability to measure and manage the full range of investment risks, and to support an investment strategy through integrated risk/ return management. The opportunity here lies in avoiding unexpected losses.

Many have acknowledged the need for better risk management in insurance. From an investment perspective, the objectives of a risk management process include:

• Calculating the risk capital required to support investment portfolios and optimize asset allocation
• Supporting investment scenarios with simulations
• Understanding the value at risk, in total and by various categories (such as asset class)
• Calculating risk-adjusted performance
• Taking account of risk-adjusted investment results in allocating overall risk capital between investment and underwriting.

Risk management systems and processes should be viewed as a journey rather than a destination. Managers need to develop them continually as they make changes to the strategic asset allocation, tactical asset allocation, and security selection levers. Risk management will help a company realize the full promise of these levers.

Efficiency and effectiveness of the investment function

Investment management is a function in which improving performance always takes priority over reducing costs. Nevertheless, even in high-performing functions, improving efficiency and effectiveness can secure benefits of up to 10 bp.

On the efficiency side, the primary opportunity lies in insourcing versus outsourcing. Since many insurance portfolios are small, lack of scale can make the function inefficient. A company might ask itself why any subscale portfolio should be managed in house if it can find external managers that will conform to its strategic, tactical, and security decisions and performance benchmarks.

A second efficiency opportunity arises from consolidating activities and purchases. Research is one area in which activities can be shared across investment units. Similarly, joint purchasing of equipment, information services, and so on can produce attractive cost reductions.

A third opportunity consists of outsourcing back-office activities such as accounting. Many global custodians and master trust providers have developed insurance investment accounting capabilities and can provide services at attractive prices.

In terms of effectiveness, the question is how best to align the organization with the investment process. Objectives for realignment include:

• Clear responsibility for results
• Greater transparency of results
• Consolidation of skills
• Faster decision making
• Improved risk management.

Exhibit 3 illustrates an organization that has been realigned with an improved investment process.

Clearly, insurers have ample opportunity to improve their investment performance. The particular levers that an insurer chooses to push and how far it chooses to push them are a function of that company’s starting point and the economic opportunities available to it. But whatever the case, the goal should be to create superior skills in the investment function and to achieve superior investment performance.